检索结果-内蒙古大学图书馆

A Novel Completion Method for Sparse Traffic Data Imputation

IEEE INTELLIGENT TRANSPORTATION SYSTEMS MAGAZINE 2025年第3期17卷 96-108页

作者： Ouyang, Renqiu Hu, Yikun Wang, Haotian Hu, Rong Yang, Wangdong Li, Kenli Hunan Univ Changsha 410082 Peoples R China Hunan Univ Comp Sci & Technol Changsha 410082 Peoples R China

Traffic data imputation is essential in smart cities and the Internet of Things (IoT). Tensor completion is an efficient method for traffic data imputation. However, these methods overlook the integration of contextual and spatial information, which are important for traffic data imputation. Hence, this study proposes a novel tensor completion method leveraging contextual and spatial information for sparse traffic data imputation (STDI). Initially, we develop a model for STDI, treating traffic data as tensors and applying tensor completion for imputing missing values. Then, to account for contextual information, we compute the contextual scores of roads and reorganize the road indices according to the scores. Additionally, we utilize the Laplacian matrix to reveal spatial information and optimize the objective function to enhance imputation accuracy. Finally, we design a parallel algorithm for STDI on GPU for efficient computation. Extensive experiments demonstrate that the proposed method is superior to existing methods.

关键词： Tensors Roads Laplace equations Imputation Trajectory Data models Sparse matrices parallel algorithms Linear programming Internet of Things

来源：评论

学校读者我要写书评

暂无评论

parallel and distributed paradigms for community detection in social networks: A methodological review

引用

EXPERT SYSTEMS WITH APPLICATIONS 2022年第0期187卷 115956-115956页

作者： Naik, Debadatta Ramesh, Dharavath Gandomi, Amir H. Gorojanam, Naveen Babu Indian Inst Technol Indian Sch Mines Dept Comp Sci & Engn Dhanbad Bihar India Univ Technol Sydney Dept Engn & Informat Technol Ultimo NSW Australia

Community detection in social networks is the process of identifying the cohesive groups of similar nodes. Detection of these groups can be helpful in many applications, such as finding networks of protein interaction in biological networks, finding the users of similar mind for ads and suggestions, finding a shared research field in collaborative networks, analyzing public health, future link prediction in social networks, analyzing criminology, and many more. However, with the increase in the number of profiles and content shared on social media platforms, the analysis is often time-consuming and exhaustive. In order to speed up and optimize the community detection process, parallel processing and Shared/Distributed memory techniques are widely used. Despite community detection has widespread use in social networks, no attempt has ever been made to compile and systematically discuss research efforts on the emerging subject of identifying parallel and distributed methods for community detection in social networks. Most of the surveys described the serial algorithms used for community detection. Our survey work comes under the scope of new design techniques, exciting or novel applications, components or standards, and applications of an educational, transactional, and co-operational nature. This paper accommodates and presents a systematic literature review with state-of-the-art research on the application of parallel processing and Shared/Distributed techniques to determine communities for social network analysis. Advanced search strategy has been performed on several digital libraries for extracting several studies for the review. The systematic search landed in finding 3220 studies, among which 65 relevant studies are selected after conducting various screening phases for further review. The application of parallel computing, shared memory, and distributed memory on the existing community detection methodologies have been discussed thoroughly. More specifically,

关键词： Social Networks Community Detection parallel algorithms Distributed algorithms

来源：评论

学校读者我要写书评

暂无评论

Enhancing Scalability and Performance in Influence Maximization with Optimized parallel Processing

Enhancing Scalability and Performance in Influence Maximizat...

引用

2024 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis, SC Workshops 2024

作者： Wu, Hanjiang Xu, Huan Park, Joongun Tithi, Jesmin Jahan Checconi, Fabio Wolfson-Pou, Jordi Petrini, Fabrizio Krishna, Tushar Georgia Institute of Technology United States Intel Corporation United States

ISBN: (纸本)9798350355543

Influence Maximization (IM) is vital in viral marketing and biological network analysis for identifying key in-fluencers. Given its NP-hard nature, approximate solutions are employed. This paper addresses scalability challenges in scale-out shared memory system by focusing on the state-of-the-art Influence Maximization via Martingales (IMM) benchmark. To enhance the work efficiency of the current IMM implementation, we propose EFFICIENTIMM with key strategies, including new parallelization scheme, NUMA-aware memory usage, dynamic load balancing and fine-grained adaptive data structures. Benchmarking on a 128-core CPU system with 8 NUMA nodes, EFFICI-ENTIMM demonstrated significant performance improvements, achieving an average 5.9x speedup over Ripples across 8 diverse SNAP datasets, when compared to the best execution times of the original Ripples framework. Additionally, on the Youtube graph, EFFICIENTIMM demonstrates a better memory access pattern with 357.4x reduction in L1+L2 cache misses as compared to Ripples. © 2024 IEEE.

关键词： data mining influence maximization parallel algorithms shared-memory social networks

来源：评论

学校读者我要写书评

暂无评论

Batch Learning SDDP for Long-Term Hydrothermal Planning

引用

IEEE TRANSACTIONS ON POWER SYSTEMS 2024年第1期39卷 614-627页

作者： Avila, Daniel Papavasiliou, Anthony Lohndorf, Nils Catholic Univ Louvain CORE B-1348 Louvain La Neuve Belgium Natl Tech Univ Athens Elect & Comp Engn Zografos 15780 Greece Univ Luxembourg L-4365 Esch Sur Alzette Luxembourg

We consider the stochastic dual dynamic programming (SDDP) algorithm - a widely employed algorithm applied to multistage stochastic programming - and propose a variant using experience replay - a batch learning technique from reinforcement learning. To connect SDDP with reinforcement learning, we cast SDDPas aQ-learning algorithm and describe its application in both risk-neutral and risk-averse settings. We demonstrate the superiority of the algorithm over conventional SDDP by benchmarking it against PSR's SDDP software using a large-scale instance of the long-term planning problem of inter-connected hydropower plants in Colombia. We find that SDDP with batch learning is able to produce tighter optimality gaps in a shorter amount of time than conventional SDDP. We also find that batch learning improves the parallel efficiency of SDDP backward passes.

关键词： Dynamic programming hydroelectric-thermal power generation parallel algorithms SDDP stochastic optimal control

来源：评论

学校读者我要写书评

暂无评论

Large-Scale K-Clustering

引用

ACM TRANSACTIONS ON KNOWLEDGE DISCOVERY FROM DATA 2024年第9期18卷 1-23页

作者： Voevodski, Konstan tin Google Inc Mountain View CA 94043 USA

Large-scale learning algorithms are essential for modern data collections that may have billions of data points. Here, we study the design of parallel k-clustering algorithms, which include the k-median, k-medoids, and k-means clustering problems. We design efficient parallel algorithms for these problems and prove that they still compute constant-factor approximations to the optimal solution for stable clustering instances. In addition to our theoretic results, we present computational experiments that show that our k-median and k-means algorithms work well in practice-we are able to find better clusterings than state-of-the-art coreset constructions using samples of the same size.

关键词： parallel algorithms large scale learning unsupervised learning clustering k-median k-medoids k-means

来源：评论

学校读者我要写书评

暂无评论

A parallel algorithm for the concurrent atomistic-continuum methodology

引用

JOURNAL OF COMPUTATIONAL PHYSICS 2022年 463卷 1页

作者： Diaz, Adrian Gu, Boyang Li, Yang Plimpton, Steven J. McDowell, David L. Chen, Youping Univ Florida Dept Mech & Aerosp Engn Gainesville FL 32611 USA Sandia Natl Labs Albuquerque NM 87185 USA Georgia Inst Technol Woodruff Sch Mech Engn Atlanta GA 30332 USA Georgia Inst Technol Sch Mat Sci & Engn Atlanta GA 30332 USA

In this work we present a parallel algorithm for the Concurrent Atomistic Continuum (CAC) formulation that can be integrated into existing molecular dynamics codes. The CAC methodology is briefly introduced and its parallel implementation in LAMMPS is detailed and then demonstrated through benchmarks that compare CAC simulation results with corresponding all-MD (molecular dynamics) results. The parallel efficiency of the algorithm is demonstrated when simulating systems represented by both atoms and finite elements. The verification benchmarks include dynamic crack propagation and branching in a Si single crystal, wave propagation and scattering in a Si phononic crystal, and phonon transport through the phase interface in a PbTe/PbSe heteroepitaxial system. In each of these benchmarks the CAC algorithm is shown to be in good agreement with MD-only models. This parallel CAC algorithm thus offers one of the first scalable multiscale material simulation methodologies that relies solely on atomic-interaction models. (c) 2022 Published by Elsevier Inc.

关键词： Concurrent atomistic-continuum method Molecular dynamics Nonequilibrium processes parallel algorithms LAMMPS

来源：评论

学校读者我要写书评

暂无评论

Using parallel Branch-and-Bound Method to Accelerate Solving Mixed Integer Linear Programming 7

Using Parallel Branch-and-Bound Method to Accelerate Solving...

引用

7th World Conference on Computing and Communication Technologies, WCCCT 2024

作者： Lei, Yang Song, Wentao Wang, Zhilei Wang, Yu Gu, Naijie University of Science and Technology of China Hefei China

ISBN: (纸本)9798350372328

The Branch-and-Bound(B&B) algorithm is an effective method for solving the Mixed Integer Linear Programming (MILP) problem. Its performance significantly impacts the overall performance of the MILP solver Additionally, parallelizing the B&B algorithm can further speed up solver efficiency. In this paper, a novel B&B parallel algorithm is proposed to accelerate the solution process of MILP problems. A Master-Worker parallel structure is employed in this algorithm, fully leverages the serial performance of the solver. To minimize overhead, only necessary information is shared between the Master and Worker to utilizing shared memory communication. During the search of the B&B tree, the Master distributes problem nodes to Workers. Workers calculate problem nodes to update the information and subsequently return a sequence of new sub-problem nodes to the Master. Once optimality is achieved or the Master has no more problem nodes, the solution process is concluded. This research demonstrates that the proposed B&B parallel algorithm exhibits excellent acceleration performance. Research indicate that the B&B parallel algorithm presented in this study exhibits a substantial speedup factor of 1.53 for instances with over 10,000 nodes within the MIPLIB2017 benchmark. Furthermore, it demonstrates an even more impressive speedup factor of up to 2.53 for instances with more than 60,000 nodes, thereby underscoring excellent performance. It is noteworthy that the relationship between the speedup factor and the number of threads is predominantly linear. © 2024 IEEE.

关键词： parallel algorithms

来源：评论

学校读者我要写书评

暂无评论

Generative Adversarial Construction of parallel Portfolios

引用

IEEE TRANSACTIONS ON CYBERNETICS 2022年第2期52卷 784-795页

作者： Liu, Shengcai Tang, Ke Yao, Xin Univ Sci & Technol China Sch Comp Sci & Technol Hefei 230026 Peoples R China Southern Univ Sci & Technol Dept Comp Sci & Engn Guangdong Prov Key Lab Brain Inspired Intelligent Shenzhen 518055 Peoples R China

Since automatic algorithm configuration methods have been very effective, recently there is increasing research interest in utilizing them for automatic solver construction, resulting in several notable approaches. For these approaches, a basic assumption is that the given training set could sufficiently represent the target use cases such that the constructed solvers can generalize well. However, such an assumption does not always hold in practice since in some cases, we might only have scarce and biased training data. This article studies effective construction approaches for the parallel algorithm portfolios that are less affected in these cases. Unlike previous approaches, the proposed approach simultaneously considers instance generation and portfolio construction in an adversarial process, in which the aim of the former is to generate instances that are challenging for the current portfolio, while the aim of the latter is to find a new component solver for the portfolio to better solve the newly generated instances. Applied to two widely studied problem domains, that is, the Boolean satisfiability problems (SAT) and the traveling salesman problems (TSPs), the proposed approach identified parallel portfolios with much better generalization than the ones generated by the existing approaches when the training data were scarce and biased. Moreover, it was further demonstrated that the generated portfolios could even rival the state-of-the-art manually designed parallel solvers.

关键词： Portfolios Training Training data Cybernetics parallel algorithms Tuning Computer science Automatic portfolio construction (APC) generative adversarial approach parallel algorithm portfolio parameter tuning

来源：评论

学校读者我要写书评

暂无评论

A parallel Algorithm Template for Updating Single-Source Shortest Paths in Large-Scale Dynamic Networks

引用

IEEE TRANSACTIONS ON parallel AND DISTRIBUTED SYSTEMS 2022年第4期33卷 929-940页

作者： Khanda, Arindam Srinivasan, Sriram Bhowmick, Sanjukta Norris, Boyana Das, Sajal K. Missouri Univ Sci & Technol Dept Comp Sci Rolla MO 65409 USA Virginia Commonwealth Univ Dept Radiat Oncol Richmond VA 23284 USA Univ North Texas Dept Comp Sci & Engn Denton TX 76201 USA Univ Oregon Dept Comp & Informat Sci Eugene OR 97403 USA

The Single Source Shortest Path (SSSP) problem is a classic graph theory problem that arises frequently in various practical scenarios;hence, many parallel algorithms have been developed to solve it. However, these algorithms operate on static graphs, whereas many real-world problems are best modeled as dynamic networks, where the structure of the network changes with time. This gap between the dynamic graph modeling and the assumed static graph model in the conventional SSSP algorithms motivates this work. We present a novel parallel algorithmic framework for updating the SSSP in large-scale dynamic networks and implement it on the shared-memory and GPU platforms. The basic idea is to identify the portion of the network affected by the changes and update the information in a rooted tree data structure that stores the edges of the network that are most relevant to the analysis. Extensive experimental evaluations on real-world and synthetic networks demonstrate that our proposed parallel updating algorithm is scalable and, in most cases, requires significantly less execution time than the state-of-the-art recomputing-from-scratch algorithms.

关键词： Heuristic algorithms Graphics processing units parallel algorithms Synchronization Multicore processing Complexity theory Wireless sensor networks Dynamic networks single source shortest path (SSSP) shared-memory parallel algorithm GPU implementation

来源：评论

学校读者我要写书评

暂无评论

Accelerating domain propagation: An efficient GPU-parallel algorithm over sparse matrices

引用

parallel COMPUTING 2022年 109卷

作者： Sofranac, Boro Gleixner, Ambros Pokutta, Sebastian Berlin Insitute Technol Str 17 Juni 135 D-10623 Berlin Germany Zuse Inst Berlin Takustr 7 D-14195 Berlin Germany HTW Berlin Treskowallee 8 D-10318 Berlin Germany

center dot Currently, domain propagation in state-of-the-art MIP solvers is single thread only. center dot The paper presents a novel, efficient GPU algorithm to perform domain propagation. center dot Challenges are dynamic algorithmic behavior, dependency structures, sparsity patterns. center dot The algorithm is capable of running entirely on the GPU with no CPU involvement. center dot We achieve speed-ups of around 10x to 20x, up to 180x on favorably-large instances.

关键词： Mixed integer linear programming MIP GPU Domain propagation Bound tightening parallel algorithms

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：